The neural system underlying conditioned taste aversion (CTA) is, like all systems underlying perceptual learning, highly complex-learning-relevant regions, most notably including basolateral amygdala (BLA) and gustatory cortex (GC), reciprocally innervate one another and then feedback on their brainstem sources of information. While most research to date has focused on the influence of each region separately rather than on their influence on each other, my lab's work is a direct investigation of inter- neuronal and inter-regional interactions during taste learning. This work has identified learning- dependent coupling between BLA and GC taste responses, each of which shifts at ~ 0.2 sec and ~0.6- 0.8 sec, and has gone on to show that CTA causes changes in GC firing that are restricted to the period following that second shift-firing that tracks taste palatability across both learning and extinction. BLA activity has been shown to be vital for this emergence of palatability-related firing, suggesting-but not proving-that BLA-GC axons are particularly vital for learning. The experiments proposed here will greatly enrich this foundational understanding of how amygdala and cortex work together during perceptual learning. We will acutely and selectively inhibit firing in neurons that project to GC from BLA, and test whether this perturbation: 1) is sufficient to block CTA learning; and 2) affects learning- and behavior-linked aspects of GC population responses appropriately (and in lock-step with perception). We will then do the inverse experiment, testing whether the impact of BLA-GC axons on learning-related population activity is itself modulated by activity in the GC-BLA pathway. Finally, we will examine the degree to which this amygdalo-cortical cooperation, and the coherent population activity that it engenders, is a function of cholinergic input from the nucleus basalis-input that is known to be important for learning, and that has been specifically proposed to mediate coherent firing in neuronal ensembles-by selectively perturbing this modulatory input (using ChAT::cre+ rats) during CTA training (coincident with recordings). Together, these experiments will directly test an uniquely systems-level view of perceptual learning, in the process revealing the neural mechanisms of a more general experiential phenomenon that affects (sometimes adversely) all mammals including humans.